Hybrid speciation in Amazonian manakins?

Rather than give a long introduction to hybrid speciation, I refer you to a recent post I did on diploid hybrid speciation in the Galápagos finches; just have a look at the introduction, which talks about the commonness of hybrid speciation in plants (via polyploidy) and its rarity in animals.  The Galápagos finches may be a case of diploid speciation following hybridization, but it’s not clear, for they don’t get a chance to mate with one of the two ancestral species (i.e., we don’t know if it’s reproductively isolated from one parent), and it’s also likely, I think, that this small-population hybrid species won’t last very long, but will be “mated to death” by its one sympatric parent.

But there’s a new report of a hybrid species in manakins from the Amazon, and this one looks pretty good. Published in PNAS (reference below, free access, and pdf here), its first author is Anfredo Barerra-Guzmán, and the last author is Jason Weir, a postdoc in Trevor Price’s lab who was next door to me for a few years.

The paper is long and complex, with some genetic details that aren’t necessary for general readers, but I’ll summarize the results as succinctly as I can.

In the Brazilian Amazon, at the headwaters of the Cururu-ri River in Pará state, exist three species of manakins: Lepidothrix vilasboasi, L. nattereri, and L. iris, which has two subspecies.  L. vilasbosi, which I’ll call LV, lives in an area intermediate between L. nattereri (“LN”) and L. iris (“LI”); all are separated by rivers, which for many Amazonian birds constitute geographic barriers because birds simply don’t like to fly over water.  Here are their geographic distributions from the paper:

You can see that LV, the putative “hybrid species” lies in between its two parents LN and LI (two subspecies). When LV was first found in 1957, was thought to be simply a “hybrid zone”: an area of hybridization between the two parental species.

The three species differ most markedly in the color of their head crowns. As the paper notes:

The crown patch in L. iris is iridescent and varies from brilliant white (its usual look, which is very similar to Lnattereri) to blue or purple, depending on the angle of light. Males of the two subspecies of Liris distributed on either side of the Xingu River are almost identical in plumage, with Liris iris possessing a thin green strip between the upper mandible and the crown patch and with the crown patch extending all the way to the mandible in Liris eucephala. Females (not shown) appear like males but lack the contrasting crown and rump patches and do not differ appreciably among species.

In contrast, LV has a yellow crown patch, very distinct from that of the two parental species (see below).

Here’s LV (the “golden-crowned manakin”) with its yellow crown:

And LN (the “snow-capped manakin”) with its white crown:

Well, what is the evidence that LV was formed after hybridization between LN and LI? That rests on genetics. The genome of LV is a mixture of genes from the two parental species, and attempts to reconstruct the ancestry of these birds, using both standard phylogenetics and “coalescent” simulations, show that it’s far more likely that LV came from mixed genes from LN and LI than that it branched off from one or the other species (the conventional “branching” scenario for speciation). Between 15% and 38% of the LV genome comes from LN and 62%-85% of the genome from LI, depending on what method is used to do the calculation.

One problem is that LV could still represent an isolated hybrid population, or a hybrid zone, between the two species. The authors say this is unlikely because pure F1 hybrids of the two species (that is, first-generation offspring), would have 50% of the genes from each species rather than this skewed distribution.  But that would be the case only if the F1 hybrids are sterile so that they couldn’t form a “hybrid swarm” by further reproduction. It’s still possible that we have here a hybrid zone or swarm in which members are not reproductively isolated from the two parental species, so that LV is not really a “species”. The crucial evidence is whether LV is reproductively isolated from LN and LI (note: it’s already geographically isolated, but that doesn’t count as a reproductive barrier in the “biological species concept”, for in that case any geographically isolated population would be a species).

So is LV reproductively isolated from the two parents? We don’t know for sure, for they don’t coexist in the same area. Barrera-Guzmán et al. suspect that LV would be isolated—by its differently colored crown. Crown colors are used by birds as ways to identify mates from their of own species, and to discriminate against the wrong ones, so the authors suppose that LV males would not be recognized as appropriate mates by females from LN and LI because of LV’s novel yellow crown. Likewise, it’s possible that LV females wouldn’t recognize the iridescent crowns of its parental species, so there would be reproductive isolation on all fronts.

The interesting thing about the LV crown is its novel yellow color. The authors found that the iridescences of the LN and LI crowns is caused by different structural properties of their feathers, and that the LV crowns are intermediate between these two sets of properties. The interesting thing is that if you make a pure F1 hybrid between LN and LI, you get a dull crown with the same intermediate structural properties of the putative hybrid LV, but it’s not yellow!

The authors then speculate—reasonably in my view—that what happened in this scenario was that the two parental species hybridized, producing a population that lives in the forest (where LV does) but had a dull crown. Sexual and/or natural selection then caused carotenoid-retaining structures to evolve in the hybrid population so that males would be visible to females in the forest. (Female preferences for yellow would, by sexual selection, go along with this.) Thus, hybridization itself wasn’t sufficient for speciation: the reproductive isolation would have to have involved some post-hybridization selection.

The upshot: I think the authors have a good case for having uncovered a rare diploid hybrid species of bird. It’s not an absolutely compelling case, as it’s still possible we have a hybrid swarm that is not a new species and is not reproductively isolated from the parents. If they could demonstrate strong mate discrimination among species based on crown color (or anything else), the data would be more convincing. But I still think they’ve uncovered a possible (even probable) case, which, with some follow-up work, could represent one of just a handful of diploid hybrid species in animals.

Some people may say, “well, this is a new evolutionary paradigm, for speciation is supposed to occur by branching of lineages, not by anastomosis (fusion) of separate branches into one.” And yes, if this is the way species normally formed, it would mandate a pretty big revision of Darwinian theory. But the evidence is that this is not the way most diploid species form, for if it did, phylogenetic analysis would not resolve any bits of the tree of life: we’d just get a spaghetti-like mess. In Drosophila, the group I know best, this is clearly not the case: branching rather than fusion is the rule. Although there is more leakage of genes between animal species than we suspected two decades ago, that is “introgresssion”, not speciation. We still see little evidence that the formation of new species in animals occurs by the fusion of already-existing branches (species).


Barrera-Guzmán, A. O., A. Aleixo, M. D. Shawkey, and J. T. Weir. 2018. Hybrid speciation leads to novel male secondary sexual ornamentation of an Amazonian bird. Proc. Nat. Acad. Sci. USA 115:E218-225. Published ahead of print December 26, 2017doi:10.1073/pnas.1717319115


  1. MJC
    Posted January 14, 2018 at 11:32 am | Permalink

    I’m not qualified to comment as invited, but I still found this interesting. Thank you.

    • Glenda Palmer
      Posted January 14, 2018 at 11:36 am | Permalink


  2. Roberto Munguia
    Posted January 14, 2018 at 11:36 am | Permalink

    Quanta magazine published a post few months ago about rethinking the relative importance of hybridization in animal evolution in the light of the recent genomic analyses.


  3. Charles Sawicki
    Posted January 14, 2018 at 11:36 am | Permalink

    Interesting case!

  4. Mark Sturtevant
    Posted January 14, 2018 at 12:00 pm | Permalink

    Very interesting! Not yet a known ‘good species’, perhaps, but still one to keep an eye on.

  5. ThyroidPlanet
    Posted January 14, 2018 at 12:20 pm | Permalink


  6. nicky
    Posted January 14, 2018 at 12:29 pm | Permalink

    Interesting, but slightly over my head. I\ll have to peruse it more thoroughly. (that may take some time, days or even weeks, hence I comment now)’
    If these rivers are an obstacle to birds, how much more they would be to, say, Marmosets or other monkeys (thinking of van Roosmalen’s many new monkeys).

  7. Edward Hessler
    Posted January 14, 2018 at 1:07 pm | Permalink

    There are several reasons I read WEIT. It is this kind of reporting that is near the top of that list, the careful explanations and caveats about the research. I especially appreciate the emphasis on evidence–the more the better (I suppose to a point.) and its consilience.

    I could, of course, read the original paper but I’d never come away with the understanding provided here.

    I, for one, wouldn’t mind some comments when/if you have the time and interest on introgression and its importance in evolution and what it says about evolutionary theory.

    And my goodness what lovely, lovely birdies. It reminds me that there is a lot going on out there!


  8. Pliny the in Between
    Posted January 14, 2018 at 1:20 pm | Permalink

    90% of why I follow this site is because of posts like this one. I rarely comment on the science posts but these are the ones I always read in detail.

    • Paul S
      Posted January 14, 2018 at 4:16 pm | Permalink

      In detail and multiple times since my lack of knowledge requires research.

  9. Posted January 14, 2018 at 3:40 pm | Permalink

    Thank you for the thoughtful analysis of this paper. Putting these results in context is especially helpful in contrast to the exaggerated speculations on display in some popularized treatments of research publications.

  10. Julian C
    Posted January 14, 2018 at 3:59 pm | Permalink

    Interesting. But, same as Nicky above, I will have to re-read carefully before I can come even close to making a cogent comment…

  11. Neil Taylor
    Posted January 14, 2018 at 4:49 pm | Permalink

    Fascinating. Thank you.

    Is there an agreed threshold for reproductive isolation? Must the off-spring be infertile?

    My understanding is quite often two species will not mate with each other as they do not recognize each other as the same species, but if they were artificially inseminated or left til the most promiscuous members just tried it as they do with everyone else the result could be viable.

    Are such cases sufficiently reproductively isolated to count as two species?

    • Brian salkas
      Posted January 14, 2018 at 11:37 pm | Permalink

      This is the age-old species-problem here. If two organisms are capable of reproducing with each other but do not in nature, then are they really two species?
      Another example could be two variations of the same “species” of flower but where they both are different colors therefor attracting different insects that only pollinate one or the other color flower, but in a lab, one could artificially pollinate them with eachother.
      Whether or not one would consider these to be two different species would depend on many other factors (morphological differences, genetic differences, viability and fertility of the offspring just to name a few) and may be a bit of a judgment call. Jerry has written about this a few times here on this site, I will also recommend the relevant bits in the book Why Evolution Is True for a succinct yet sufficient non-technical summary of different criteria for what a species actually is and how different species are delineated.
      My non expert answer here would be that there is some grey area when it comes to delineating species. Well qualified scientists can, and do, disagree about the details, and it depends how much weight one gives to certain criteria.

      • ThyroidPlanet
        Posted January 15, 2018 at 1:25 am | Permalink

        Can you give an example of that?

    • Pierluigi Ballabeni
      Posted January 15, 2018 at 7:55 am | Permalink

      As far I understand, if individuals from two species do not mate with each other in nature in spite of their potentially being interfertile, we have a case of prezygotic (i.e. before fertilisation) isolation. The species are actually isolated.

  12. Posted January 14, 2018 at 7:51 pm | Permalink

    Why does hybrid speciation happen so much more often in plants?

    • Posted January 15, 2018 at 10:47 am | Permalink

      The most important reason is that plants can reproduce vegetatively or asexually in various ways. The easiest way to see this is to consider what occurs in animals, most of which cannot reproduce vegetatively/asexually. When a horse and donkey cross to produce a mule, the mule’s gametes are almost all defective due to mismathcing of the two parental chromosome sets during meiosis (part of gamete formation). So, when an individual mule dies, that’s the end of its genetic line. But a hybrid plant with mismatched chromosomes can reproduce vegegtatively, and thus survive indefinitely. Eventually, a mutation producing a doubling of the chromosomal complement may occur. (Such mutations, though rare, do occur.) Once the chromosome complement is doubled, normal meiosis can take place, each parental set now having a matching pair: the chromosome complement goes from AB (A and B representing the two parental species) to AABB, so A pairs with A and B pairs with B. So, you now have a sexual species reproductively isolated from both parental forms.

      Many such cases are known in plants. An especially well known case is the formation of Spartina anglica from the parental species S. alterniflora and S. maritima. The sterile vegetatively reproducing hybrid between the latter two, known as S. townsendi, gave rise to anglica via a chromosomal doubling. This all occurred during the 19th century, and is probably the best documented case of speciation within a (long) human lifetime in nature.

      It may also be the case that plants’ simpler developmental systems allow more initial hybrids to survive, so that there are more candidates to enter the vegetative phase of the process.

      • Posted January 15, 2018 at 6:01 pm | Permalink

        Thanks for the thorough explanation!

  13. W.Benson
    Posted January 14, 2018 at 8:44 pm | Permalink

    I’m not so convinced. These small fruit-eating live in the forest interior and rivers, even narrow rivers a few 10s of meters wide, can be a strong barrier to gene flow. Would it not be simpler just to assume that after the ancestral species spread across forests of the southern Amazon, all populations occasionally produced mutations for yellowish caps, whereas between the R. Tapajos and Jamanxim yellow males were more fit and yellow caps evolved, perhaps as an indicator of nutrition for female choice for males good at finding berries and who would tend to produce offspring with similar abilities. There is conspicuous environmental gradient from the eastern Amazon, where it is drier and more hilly, to the West where it is wetter with deep soils. The importance of different foods, predators, parasites and diseases probably vary in parallel. It is not strange that under intermediate ecological conditions and an intermediate geographic location, isolated by rivers, L. villaboasi might exhibit a unique phenotype associated with a intermediate genetic composition.
    I will now read the report to see if the authors and the complete data can convince me.

  14. largeswope
    Posted January 15, 2018 at 7:55 am | Permalink


  15. jaxkayaker
    Posted January 15, 2018 at 11:27 am | Permalink

    Very interesting, Jerry. Many thanks for posting. I’m sending it to my M.S. advisor.

    In the penultimate sentence, there’s a typo: “introgression” has one s too many.

  16. Posted January 15, 2018 at 6:08 pm | Permalink

    This is really interesting to me. The evolution of the distinctive yellow crown suggests the hybrid was under selection to not breed with its neighboring parent species. This is also a common occurrence in other closely-related species pairs in zones of contact.

  17. Liz
    Posted January 17, 2018 at 2:16 pm | Permalink

    “…using both standard phylogenetics and “coalescent” simulations, show that it’s far more likely that LV came from mixed genes from LN and LI than that it branched off from one or the other species (the conventional “branching” scenario for speciation).”

    1. What is an example of the conventional branching? A donkey and a horse would be like these birds, correct, and not conventional branching?

    “It’s still possible that we have here a hybrid zone or swarm in which members are not reproductively isolated from the two parental species, so that LV is not really a “species”.”

    2. Does this mean that LV would be a swarm if they are able to reproduce with each other to make fertile offspring (or any offspring), or does the swarm mean that they would exist and be able to reproduce with both parent species?

    “Sexual and/or natural selection then caused carotenoid-retaining structures to evolve in the hybrid population so that males would be visible to females in the forest.”

    3. Is this saying that the yellow hair forming is the defining factor in making this hybrid a species of its own? Or was it its own species already? Either way, is it still able to mate with the parent species or do the LVs just mate by themselves now creating fertile offspring?

    4. What exactly is diploid hybrid speciation by definition? I had trouble finding a definition or piecing it together from this article.

    Thanks so much!

  18. Posted February 4, 2018 at 10:08 am | Permalink

    Interesting! It seems that, with regard to speciation, we should never say “never”.

%d bloggers like this: